Objectives: The opioid epidemic and the associated deaths have increased the availability of increased-risk donor organs. Here, we assessed factors associated with increased-risk donor liver transplant and determined their impact on survival and response to direct-acting antivirals.
Patients and Methods: We analyzed anti-hepatitis C virus-positive deceased-donor liver transplant recipients from August 2013 through December 2017. We compared recipient and donor clinical and virologic features, response to direct-acting antivirals, and graft and patient survival rates in increased-risk versus tradi-tional or non-increased risk donor organ transplants.
Results: Of 153 transplant recipients, 89 (58%) were anti-hepatitis C virus positive, with 42/89 receiving increased-risk donor livers (mean age 62 years, 1 female, 80% white, and 60% with hepatoma). On univariable analysis, receipt of increased-risk donor liver was associated with simultaneous liver-kidney transplant, lower Model for End-Stage Liver Disease score, hepatitis C virus RNA positivity, pretransplant direct-acting antiviral nonresponse, and younger donor age. On multivariable analysis, only donor age and Model for End-Stage Liver Disease score were associated with increased-risk donor transplant. Among increased-risk donors, 12 (29%) were hepatitis C virus RNA positive, including one who was anti-hepatitis C virus antibody negative. Among recipients, 62 were hepatitis C virus RNA positive (35 with increased-risk livers), with 50 recipients (81%) having genotype 1. Posttransplant, recipient genotype changed in 6 and was mixed in 4 recipients. Of 55 recipients treated with direct-acting antivirals, 54 (98%) achieved viral clearance. Overall 1-year graft and patient survival was 93%.
Conclusions: Increased-risk donor organs provided high levels of utility in liver transplant recipients who were anti-HCV positive, showing optimal graft and patient survival. Increased-risk donors were younger and preferably transplanted in hepatitis C virus RNA-positive recipients with lower Model for End-Stage Liver Disease score. Posttransplant direct-acting antiviral therapy was highly efficacious irrespective of pretransplant recipient and donor virologic status.
Key words : Donor organ, Graft survival, Opioid epidemic, Outcome research
The association between organ transplant and posttransplant hepatitis was first described almost 4 decades ago.1 The discovery of hepatitis C virus (HCV) led to the realization that most cases of non-A, non-B posttransplant hepatitis were due to HCV.2 As a result, routine donor testing for HCV and informed consent from potential recipients of seropositive organs were instituted. A number of studies have indicated that transplant of HCV-positive livers into HCV-positive recipients did not adversely affect graft and patient outcomes.3-5 The introduction of anti-HCV antibody assay and later nucleic acid testing diminished the risk of viral transmission by narrowing the window between exposure and diagnosis.6 The incidence of window-period HCV infection declined 10-fold once nucleic acid testing was implemented compared with the era of antibody testing alone.7
In July 2013, the United States Public Health Service (PHS) introduced updated guidelines for reducing transmission of human immunodeficiency virus, hepatitis B virus (HBV), and HCV through organ transplant.8 Donors were deemed to be at increased risk for HCV transmission if they met 1 or more of the 12 criteria for risk assessment. The opioid epidemic and the associated surge in intravenous drug overdoses have increased the number of organs from increased-risk donors (IRD).9 A recent review of the Organ Procurement and Transplantation Network database indicated that the use of such organs may not result in inferior graft or patient survival.10 Indeed, it could be argued that the use of HCV-positive organs and resultant lower discard rate would enhance overall survival by reducing wait list mortality.
Concurrent with the application of the PHS guidelines, HCV therapy has witnessed a paradigm shift that has resulted from the availability of well-tolerated and highly effective direct-acting antivirals (DAAs).11 A wider use of IRD organs may thus be recommended; however, it would be prudent to evaluate this strategy before it is universally adopted.
The aims of our study were to assess factors associated with IRD liver transplants and to determine the impact of the use of such organs on response to DAAs and on graft and patient survival.
Materials and Methods
The Veterans Affairs Pittsburgh Healthcare System is the largest provider of liver transplants within the Veterans Health Administration (www.srtr.org; accessed June 2018). Organ recipients are routinely followed to ensure appropriate management of immunosuppression and of possible posttransplant complications. In this study, we conducted a retrospective analysis of liver allograft recipients who were seropositive for HCV. The study was approved by our institutional review board and conformed to the ethical guidelines of the 1975 Helsinki Declaration. Recipients who underwent liver transplant between August 1, 2013 and December 31, 2017 were identified by a search of the transplant database. This period was selected in view of the release of revised PHS donor guidelines in July 2013.8 Incidentally, the US Food and Drug Administration approved second-generation DAAs beginning in November 2013.12 The study cohort was categorized into those who received IRD allografts versus those who received non-IRD allografts.
We examined recipient demographics and disease severity as determined by Child class and Model for End-Stage Liver Disease (MELD) score. Recipients were deemed to have hepatocellular carcinoma (HCC) if they were diagnosed to have HCC by imaging and/or by histology pretransplant or by pathologic assessment of the liver explant. Simultaneous liver-kidney (SLK) transplant and immunosuppressive regimens were noted. We evaluated donor demographics and viral studies and noted whether a donor was deemed to be an IRD according to the PHS criteria. Hepatitis C serology and HCV RNA levels and genotypes were noted before and after liver transplant. We also evaluated therapies received for HCV and the response to those therapies. Patient and graft survival rates were noted; however, primary graft nonfunction was disregarded for the purpose of survival analysis.
We used descriptive statistics to summarize recipient and donor features. Categorical variables were reported as proportions and compared by chi-square test. Continuous variables were noted as means with standard deviation and range; comparisons across categories were done by the 2-tailed t test. To determine variables predictive of transplant with an IRD liver, we performed logistic regression analysis with the forced entry method. Factors determined to be significant on univariable analysis were entered in the model. The significance of the model was assessed by calculating R2 value, while the individual contribution of predictors was estimated with the Wald statistic. We also calculated the odds ratio for each predictive variable. Graft and patient survival rates were assessed by Kaplan-Meier analysis, and the 2 patient groups were compared by log-rank test. Statistical analyses were performed with SPSS software (SPSS: An IBM Company, version 25.0, IBM Corporation, Armonk, NY, USA).
During the 3.5-year period between August 1, 2013 and December 31, 2017, 153 patients received liver allografts. Among them, 90 (59%) were seropositive for HCV, including 1 patient coinfected with HBV. Sixty-three patients (41%) had end-stage liver disease from other causes, including nonalcoholic steato-hepatitis in 27 (18%), alcohol alone in 22 (14%), α1-antitrypsin deficiency in 4, and liver allograft failure in 5 patients.
After we excluded the coinfected patient with HBV, 89 HCV-positive patients were further studied (Table 1). Among these recipients, 42 (47%) received IRD liver allografts and 47 (53%) received organs from non-IRDs, as defined by the revised PHS guidelines.8 Recipient demographics were similar in the 2 groups. The mean age was 62 years, and only 3 of the 89 patients evaluated were younger than 50 years old. All except 1 recipient were men, and 71 (80%) were white. Fifty-three recipients (60%) harbored HCC with similar proportions in the 2 groups. Ten patients (11%) underwent SLK transplant, with a significantly higher proportion in the IRD group. The average MELD score at transplant of the entire cohort was 27 (range, 10-40), with significantly lower MELD scores in the recipients who received IRD organs compared with those who received non-IRD organs. Blood group distribution was similar in the 2 groups, with blood groups O and A constituting 47% and 36% of the recipients, respectively.
As shown in Table 1, the average age of donors was 40 years (range, 18-69 y). Increased-risk donors were significantly younger than non-IRDs, with a difference of 11 years in their mean age. Thirty-six of the 42 IRDs (86%) were younger than 40 years, compared with 18 of the 47 non-IRDs (38%). Of 42 IRDs, 27 (64%) were anti-HCV positive; among this group, 11 (41%) were HCV RNA positive. Of the 15 anti-HCV negative IRDs, one was HCV RNA positive. Thus, in total, 12 IRDs (29%) were HCV RNA positive. Eleven other donors with history of intravenous drug use were negative for both anti-HCV antibody and HCV RNA. None of the non-IRDs was anti-HCV antibody or HCV RNA positive.
Hepatitis C pretransplant
As shown in Table 2, 70/89 patients (79%) received antiviral therapy before transplant, including 36 (40%) who received DAAs and 34 (38%) who received interferon-based therapies alone. The proportions of patients who received DAA versus interferon alone were similar in the 2 groups. Among the 36 patients treated with DAA, 22 (61%) achieved a sustained virologic response (SVR) with a significantly higher proportion in the non-IRD group. None of the recipients were on antiviral therapy at the time of transplant. Overall, 62 patients (70%) were HCV RNA positive at transplant, and this group was more likely to receive an IRD organ. Among those patients, 50 (81%) had genotype 1 infection; pretransplant viral RNA levels and genotypes were similar in the 2 groups.
Hepatitis C posttransplant
As shown in Table 2, at the time of transplant, 62/89 recipients (70%) were HCV RNA positive. After we excluded 2 patients who died early, all 60 recipients who were HCV RNA positive pretransplant were noted to be viremic posttransplant, whereas none of the RNA-negative recipients showed viremia after transplant. The average viral level was 5.8 million IU posttransplant, with no significant differences between the 2 groups. Eight patients exceeded the limit of detection at 25 million IU. Hepatitis C genotype was tested in 47 of the 60 patients posttransplant, including 32 patients in the IRD group. A change in genotype was noted in 6 patients, whereas mixed genotypes were found in 4 recipients. Among the 5 patients with genotype 2 pretransplant, genotype changed to 1a in 4 patients and remained the same in 1 patient. Among the 3 patients with genotype 3 pretransplant, genotype changed to 1a in 1 patient and to genotype 2 in another patient; genotype remained unchanged in the third patient. Two patients with genotype 1a (1 with genotype 1b and 1 with genotype 2) developed mixed genotypes, and all had genotype 1a mixed with other genotypes. None of the HCV RNA-negative recipients was noted to be viremic after transplant.
Hepatitis C treatment posttransplant
As shown in Table 3, among the 62 viremic patients posttransplant, 54 were treated with DAAs. Fourteen of the 54 treated recipients (26%) had relapsed or did not respond to DAA pretransplant (with 11 in the IRD group). Of the 46 treated patients with genotype 1, 1 was treated with sofosbuvir/ribavirin, 2 with sofosbuvir/simeprevir, 1 with sofosbuvir/ribavirin/-glecaprevir/elbasvir, and 1 with sofosbuvir/-valpatasvir/ribavirin. Thirty-four patients received sofosbuvir/ledipasvir combination (18 with ribavirin). Five received glecaprevir/pibrentasvir combination (1 with ribavirin). One patient who received sofosbuvir/ledipasvir pretransplant relapsed and was treated with a sofosbuvir/-valpatasvir/voxilaprevir fixed-dose combination. There was only 1 patient with genotype 2 who received glecaprevir/pibrentasvir. There were 3 patients with genotype 3, with 1 patient treated with sofosbuvir/daclatasvir/ribavirin and 2 patients treated with sofosbuvir/valpatasvir (1 in com-bination with ribavirin).
All patients responded to therapy and achieved SVR. The only patient with genotype 4, a nonresponder to pegylated-interferon and ribavirin pretransplant, was treated with sofosbuvir/-ledipasvir/ribavirin to which he responded but subsequently relapsed. Four patients with mixed genotypes, all with genotype 1a, underwent treatment with sofosbuvir-based regimens (n = 3) or glecaprevir/pibrentasvir (n = 1); all 4 achieved SVR. All patients tolerated therapies well with no notable adverse effects.
Figure 1 and Figure 2 show survival results. Two recipients died in the IRD group, and 8 recipients died in the non-IRD group. One-year patient survival was 93% in the entire cohort, 98% in the IRD group, and 88% in the non-IRD group (not significantly different).
As shown in Table 4, on univariable analyses, features associated with transplant of IRD livers included SLK transplant, MELD score, donor age, recipient’s HCV RNA positivity, and response to DAA pretransplant. We performed logistic regression analysis to determine factors predictive of im-plantation of an IRD liver by forced entry of all those factors except pretransplant response to DAA. We excluded this factor in view of colinearity between HCV RNA positivity and response to DAA. We found MELD score and donor age to be predictive of transplant with an IRD liver.
Thus, livers considered to be of increased risk were likely to be from younger donors and were transplanted into recipients with lower MELD. Other variables that differed significantly on univariable analysis (SLK transplant and recipient HCV positivity) had no significant association with IRD transplant on multivariable analysis.
Three decades after its introduction, liver transplant remains the principal life-saving therapy for patients with advanced liver disease. Unfortunately, its application has remained hampered by the persistent gap between the demand and supply of donor organs, resulting in considerable wait-list mortality.13 The emergence of the opioid epidemic has had a devastating effect on the US adult population, resulting in an estimated 42 000 deaths in 2016.14 Recent studies have suggested that the reported fatality rate may be an underestimate, and the actual number of deaths is substantially more.15 Inci-dentally, opioid-related deaths have resulted in an increase in organ availability, although such donors are deemed to be at increased risk for transmission of HBV, HCV, and human immunodeficiency virus.8 Faced with the long wait-list time for donor livers, many transplant centers have opted to use IRD organs. An analysis of data from the Scientific Registry of Transplant Recipients has indicated that 17% of HCV-positive recipients received HCV-positive donor livers in 2015; however, the discard rate of HCV-positive livers remained almost twice as high as HCV-negative livers.16 It would seem acceptable to transplant an organ from an HCV RNA-positive donor to an HCV RNA-positive recipient. More questionable would be the decision to transplant a liver from an HCV RNA-positive donor into an HCV RNA-negative recipient with or without anti-HCV, as that would certainly lead to new-onset HCV infection. Nucleic acid testing has reduced infection risk, although it has not completely eliminated it.7 Reassuringly, posttransplant HCV infection can now be successfully treated due to the availability of DAAs; however, there is a theoretical risk of infection with multidrug resistant strains that may not be amenable to therapy.17
Our present study provided a single-center experience of liver transplant among HCV sero-positive recipients over a period that followed the implementation of the most recent PHS IRD guidelines.8 Incidentally, this has also allowed us to assess the impact of second- and third-generation DAAs. The study cohort demographics were representative of the US veteran population with advanced liver disease, but these differ from US national liver recipient demographics.13 In our cohort, almost all were men, 80% were white, and 97% were older than 50 years, whereas nationally those proportions are 65%, 71%, and 77%, respectively. A high proportion (60%) of our patients were noted to have HCC; in the US overall, 20% received liver transplants with an HCC MELD exception.13 The latter figure was likely an under-estimate, as that referred exclusively to recipients with MELD exception, whereas our data also included recipients without exception and those noted to have HCC in the explant.
In our study, among the factors that differed significantly between IRD versus non-IRD groups on univariable analyses, only recipient MELD score and donor age were predictive of an IRD liver transplant. Increased-risk donors were considerably younger and were implanted into recipients with lower MELD scores. Those findings were possibly reflective of the transplant decision-making process. In view of the high rejection and discard rate of IRD livers, it was likely that patients with lower MELD scores accepted such offers instead of waiting longer for a non-IRD organ. Conversely, recipients with higher MELD scores were more likely to have offers of traditional or non-IRD organs. Younger age of IRD was reflective of the opioid epidemic, which predominantly affects young adults. A recent study reported an increase in the annual incidence of acute HCV from 0.3 per 100 000 in 2004 to 0.7 per 100 000 in 2014.9 Injection drug use was reported by 84% of the patients in 2014. The largest increases in annual incidence rate were in the 18- to 29-year (400%) and 30- to 39-year (325%) age groups. Although HCV RNA positivity of recipients did not predict receipt of IRD livers on multivariable analysis, it differed between the 2 groups. That was an expected finding, as HCV RNA-positive recipients were more likely to receive organs with the potential, albeit small, risk of HCV transmission. In our study, donor HCV seropositivity did not correlate strongly with HCV RNA positivity as 16/27 (59%) of anti-HCV positive donors were RNA negative, whereas 1 of the 15 anti-HCV-negative IRDs was RNA positive. Possibly, anti-HCV-positive RNA-negative donors had spontaneous viral clearance or had been treated for HCV infection. One seronegative donor who was HCV RNA positive likely had recent infection and was in the window period for anti-HCV antibody.
The increasing application of DAAs was evi-denced by the fact that 40% of our recipients had received therapy pretransplant. The SVR rate of 61% among those patients was lower than the rate reported in clinical trials and closer to the rate reported among patients with decompensated cirrhosis.18,19 Many of our patients had advanced liver disease, and that may have compromised their response to treatment. Recurrent HCV infection was universal among recipients viremic at transplant, whereas none of the nonviremic recipients developed de novo HCV infection. Interestingly, many of the recipients had a change in genotype or mixed genotypes after transplant that indicated donor-derived infection. Genotype 1a appeared to be the dominant genotype, particularly among recipients with prior genotype 2, and it persisted among recipients with mixed genotypes posttransplant. Direct-acting antivirals were well tolerated and highly effective posttransplant. Only 1 patient (genotype 4 infected) relapsed after achieving an end of treatment response, whereas all others achieved SVR. We noted no impact of IRD allografts on transplant outcomes, as graft and patient survival rates were as expected with no significant difference between the 2 groups.
Limitations of our study include the single-center design and the cohort size. We were constrained by the time frame, which we selected to allow us to assess the impact of the revised PHS guidelines and of second- and third-generation DAAs. Nevertheless, we corroborated the findings of previously published studies and in addition provided more detailed information, in particular with regard to HCV genotypes.
Increased-risk donor liver allografts, which became available mostly from younger donors, provided a high level of utility in HCV seropositive recipients with optimal graft and patient survival. Posttransplant DAA therapy was highly efficacious irrespective of pretransplant recipient and donor virologic status. Our findings support the practice of transplanting IRD organs into HCV seropositive recipients, a policy that is being increasingly adopted by the transplant community. In view of the excellent response to DAAs, we recommend an expansion of this policy to HCV-negative recipients.
DOI : 10.6002/ect.2019.0065
From the 1Veterans Affairs Pittsburgh Healthcare System, the 2Division of
Gastroenterology, Hepatology, and Nutrition, and the 3Division of Transplant
Surgery, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania,
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare.
Corresponding author: Obaid S. Shaikh, VA Pittsburgh Healthcare System, University Drive C, FU #112, Pittsburgh, PA 15240, USA
Phone: +1 412 480 6109
Table 1. Demographics of Recipients and Donors in the Study Groups
Table 2. Hepatitis C Virus Pretransplant and Posttransplant Variables in the Study Groups
Table 3. Hepatitis C Treatment Posttransplant
Table 4. Variables Predictive of Increased-Risk Donor Transplant
Figure 1. Patient Survival
Figure 2. Patient Survival in Increased-Risk Donor versus non-Increased-Risk Donor